有人与无人驾驶车辆交叉口驾驶博弈模型

针对有人与无人驾驶车辆在交叉口存在冲突时的协调控制问题，引入智能网联车辆的设计思想，将交叉口存在交互行为的决策个体建模为博弈中的参与者，以冲突车辆的速度改变方案为博弈策略，构建双方的收益矩阵，而驾驶收益采用行车安全收益、行车效率收益和行车舒适性收益来计算，求解博弈模型的纳什均衡，作为双方的最优驾驶策略组合，完成交叉口多车冲突的协作优化.模型加入驾驶员类型的多样性模拟，基于Matlab对提出的算法进行验证，结果表明无人驾驶车辆会根据对方驾驶员行为调整自身的行为策略，与基于冲突表的协作算法对比，本算法的冲突消解所用时间更短，在确保安全的同时提高了冲突车辆通过路口的效率. 

A Driving Game Model for Manned and Unmanned Vehicles at Intersection

To make a coordinated control for the conflict between manned and unmanned vehicles at intersections, a coordination algorithm was designed based on game theory, regarding the manned and unmanned vehicles at intersections as participants in the game with the thought of intelligent connected vehicle. Firstly, taking the speed change scheme as the game strategy, a profit matrix was constructed for both sides. And then, driving revenue function was determined by introducing the benefits of driving safety, driving efficiency and driving comfort. The Nash equilibrium of the game model was solved as the optimal driving strategy combination of both sides to complete the collaborative optimization of multi-vehicle conflict at intersections. Finally, the diversity of driver types was introduced into the simulation model,validating the proposed algorithm with Matlab software. The simulation results show that, the unmanned vehicle can adjust its behavior strategy according to the behavior of human driver. Compared with conflict table algorithm, the coordination algorithm can reduce the time of conflict resolution obviously. It can be seen that the model not only effectively avoids the collision risk, but also improves crossing efficiency at intersections.